{
  "affected": [
    {
      "affectedData": [
        {
          "product": "LightRAG",
          "vendor": "HKUDS",
          "versions": [
            {
              "status": "affected",
              "version": "\u003c 1.5.5"
            }
          ]
        }
      ],
      "source": "security-advisories@github.com"
    }
  ],
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "LightRAG provides simple and fast retrieval-augmented generation. Prior to 1.5.5, _validated_addresses in lightrag/parser/markdown/parser.py evaluates the literal resolved address with ipaddress.is_global without consistently classifying an IPv4 address embedded in an IPv6 transition wrapper. A caller who can upload a Markdown or textpack document can supply an external image URL using NAT64 64:ff9b::/96 or an IPv4-compatible form that embeds a loopback, private, or cloud-metadata IPv4 address. On a deployment with compatible NAT64 or DNS64 routing, _download and _build_guarded_opener accept the wrapper and fetch the internal resource, whose body is then ingested. Current interpreter behavior already blocks some RFC 8215 and 6to4 forms, but the fixed guard handles all documented wrappers without becoming more permissive than the standard library. This issue is fixed in version 1.5.5."
    }
  ],
  "id": "CVE-2026-85740",
  "lastModified": "2026-09-28T20:17:11.063",
  "metrics": {
    "cvssMetricV31": [
      {
        "cvssData": {
          "attackComplexity": "HIGH",
          "attackVector": "NETWORK",
          "availabilityImpact": "NONE",
          "baseScore": 7.1,
          "baseSeverity": "HIGH",
          "confidentialityImpact": "HIGH",
          "integrityImpact": "LOW",
          "privilegesRequired": "LOW",
          "scope": "CHANGED",
          "userInteraction": "NONE",
          "vectorString": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:L/A:N",
          "version": "3.1"
        },
        "exploitabilityScore": 1.8,
        "impactScore": 4.7,
        "source": "security-advisories@github.com",
        "type": "Secondary"
      }
    ],
    "ssvcV203": [
      {
        "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
        "ssvcData": {
          "id": "CVE-2026-85740",
          "options": [
            {
              "exploitation": "poc"
            },
            {
              "automatable": "no"
            },
            {
              "technicalImpact": "partial"
            }
          ],
          "role": "CISA Coordinator",
          "timestamp": "2026-09-28T19:33:40.322379Z",
          "version": "2.0.3"
        }
      }
    ]
  },
  "published": "2026-09-22T17:17:27.520",
  "references": [
    {
      "source": "security-advisories@github.com",
      "url": "https://github.com/HKUDS/LightRAG/commit/9207e7fd10cf6df3cf26e3dd1921490b34a2f132"
    },
    {
      "source": "security-advisories@github.com",
      "url": "https://github.com/HKUDS/LightRAG/commit/a25862177a9b00e0edb870778a42b4155924de09"
    },
    {
      "source": "security-advisories@github.com",
      "url": "https://github.com/HKUDS/LightRAG/commit/c598545aa0084a427bcb19a84ade5e0ec31fa673"
    },
    {
      "source": "security-advisories@github.com",
      "url": "https://github.com/HKUDS/LightRAG/pull/3426"
    },
    {
      "source": "security-advisories@github.com",
      "url": "https://github.com/HKUDS/LightRAG/releases/tag/v1.5.5"
    },
    {
      "source": "security-advisories@github.com",
      "url": "https://github.com/HKUDS/LightRAG/security/advisories/GHSA-vv3m-f8x4-7377"
    },
    {
      "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
      "url": "https://github.com/HKUDS/LightRAG/security/advisories/GHSA-vv3m-f8x4-7377"
    }
  ],
  "sourceIdentifier": "security-advisories@github.com",
  "vulnStatus": "Deferred",
  "weaknesses": [
    {
      "description": [
        {
          "lang": "en",
          "value": "CWE-918"
        }
      ],
      "source": "security-advisories@github.com",
      "type": "Secondary"
    }
  ]
}

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "lightrag-hku"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.5.5"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-85740"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-918"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-09-22T20:40:25Z",
    "nvd_published_at": "2026-09-22T17:17:27Z",
    "severity": "HIGH"
  },
  "details": "## Summary\n\nLightRAG\u0027s native markdown parser downloads external images referenced by an uploaded markdown or textpack document. The only SSRF guard, `_validated_addresses()` in `lightrag/parser/markdown/parser.py`, resolves the image host and rejects it when the resolved IP is not `is_global`. That check is evaluated on the raw resolved address and never decodes IPv6 transition wrappers that embed an internal IPv4. Python\u0027s `ipaddress` classifies a NAT64 (`64:ff9b::/96` and the RFC 8215 `64:ff9b:1::/48` prefix), IPv4-compatible (`::a.b.c.d`), or 6to4 (`2002::/16`) literal that wraps an internal IPv4 as globally routable, so the guard passes it. On a host with NAT64/DNS64 routing the request is then delivered to the embedded internal target (loopback, RFC1918, or a cloud metadata endpoint), and the fetched body is ingested. The plain and IPv4-mapped (`::ffff:`) forms of the same internal address are correctly blocked, so this is an encoding that defeats the existing filter.\n\n## Affected component and versions\n\n- Package: `lightrag-hku` (LightRAG), the native markdown image-download path.\n- Component: `lightrag/parser/markdown/parser.py`, guard `_validated_addresses()` (the `if not (ip.is_global or ...)` check), reached from `_download()` -\u003e `_build_guarded_opener().open(req)`.\n- Enabled by default: `download_enabled = _env_bool(\"NATIVE_MD_IMAGE_DOWNLOAD_ENABLED\", True)`.\n- Affected: LightRAG `\u003c= 1.5.4` (latest release at time of report, commit `9a45b64`).\n- Precondition: the caller can upload a document (API key via `Depends(combined_auth)`); the host has NAT64/DNS64 routing for the encoded address to reach the internal endpoint.\n\n## Vulnerable code vs the guarded sibling\n\nEvaluated on the exact resolved addresses (CPython `ipaddress`):\n\n| Form | address | is_global | guard verdict |\n|---|---|---|---|\n| plain internal 127.0.0.1 | `127.0.0.1` | False | BLOCK (correct) |\n| IPv4-mapped | `::ffff:7f00:1` | False | BLOCK (correct) |\n| NAT64 64:ff9b::/96 | `64:ff9b::7f00:1` | True | PASS (bypass) |\n| NAT64 RFC8215 64:ff9b:1::/48 | `64:ff9b:1::7f00:1` | True | PASS (bypass) |\n| IPv4-compatible ::a.b.c.d | `::7f00:1` | True | PASS (bypass) |\n| 6to4 2002::/16 | `2002:7f00:1::` | True | PASS (bypass) |\n\nThe guard already blocks the plain and IPv4-mapped internal forms (the forms the author tested), proving the intent is to reject internal destinations; it never canonicalizes the other transition wrappers, so they pass.\n\n\u003e **Note \u2014 CPython version dependency of the table.** The `is_global` verdicts above depend on the interpreter. CPython gh-113171 (backported to 3.10.14 / 3.11.9 / 3.12.3+, ~2024-04) added `64:ff9b:1::/48` and `2002::/16` to `ipaddress._private_networks`, so on any current patch release those two rows return `is_global = False` and are already blocked by the stdlib before the guard is reached. The **`64:ff9b::/96` well-known NAT64 prefix and the IPv4-compatible `::/96` form bypass on every current CPython** \u2014 and `64:ff9b::/96` is exactly the prefix real DNS64/NAT64 deployments use, so the vulnerability stands. The four-row table reflects a pre-gh-113171 interpreter. The fix classifies by the embedded IPv4 and therefore closes all forms regardless of interpreter version.\n\n## Severity\n\nHigh. CWE-918 (Server-Side Request Forgery). CVSS 3.1 vector `AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:L/A:N`, base score 7.1. AC:H reflects the NAT64/DNS64 network precondition for end-to-end reach; PR:L because document upload requires the API key; S:C because the request pivots into an internal network segment; C:H for internal/metadata/secret disclosure.\n\n## Proof of concept (deployed, both directions, verbatim)\n\nEnvironment: three Docker containers on an IPv6-enabled network - an internal victim (`10.66.0.2:80`, RFC1918) serving a unique secret, a DNS64/NAT64 gateway owning `64:ff9b::/96` that forwards `[64:ff9b::0a42:0002]:80` to `10.66.0.2:80`, and an attacker host with a client-side route `64:ff9b::/96 via \u003cgateway\u003e`. This models a client on an IPv6-only / NAT64+DNS64 network (AWS/GCP IPv6-only subnets, many mobile and corporate networks). The attacker runs the EXACT released guard code (`_validated_addresses`, `_host_is_public`, `_build_guarded_opener`, the `_Guarded*Connection/Handler` classes, sliced verbatim from v1.5.4 `parser.py`, only the `logger` import shimmed) plus a verbatim reproduction of the released `_download()` scheme and host precheck. Trigger in production: an uploaded document containing `![x](http://[64:ff9b::\u003cinternal-hex\u003e]/a.png)`.\n\nRUN 1, released v1.5.4 guard (VULNERABLE):\n\n```\nattack URL = http://[64:ff9b::0a42:0002]:80/logo.png\n===== DIRECTION A: ATTACK (NAT64-wrapped internal address) =====\nguard verdict : PASSED (fetched)\nSECRET LEAKED : True\nbody          : SSRF-LIGHTRAG-INTERNAL-METADATA-9f3a2b7c\n===== control: plain internal address (must be BLOCKED) =====\nverdict: BLOCKED at guard: non-public host blocked\n===== DIRECTION B: PUBLIC address (must still be FETCHED) =====\nguard verdict : PASSED (fetched)\npublic body   : PUBLIC-OK-BODY\n===== SUMMARY =====\nattack_leaked_internal_secret = True\nplain_internal_blocked        = True\npublic_still_fetched          = True\n```\n\nRUN 2, one-line-class decode fix applied (same inputs, same network):\n\n```\n===== DIRECTION A: ATTACK (NAT64-wrapped internal address) =====\nguard verdict : BLOCKED at guard: non-public host blocked\nSECRET LEAKED : False\n===== control: plain internal address (must be BLOCKED) =====\nverdict: BLOCKED at guard: non-public host blocked\n===== DIRECTION B: PUBLIC address (must still be FETCHED) =====\nguard verdict : PASSED (fetched)\npublic body   : PUBLIC-OK-BODY\n===== SUMMARY =====\nattack_leaked_internal_secret = False\nplain_internal_blocked        = True\npublic_still_fetched          = True\n```\n\nFix-correctness matrix (isolated): genuine global IPv6 `2606:4700:4700::1111` and `2001:4860:4860::8888` -\u003e ALLOW (no false positive); NAT64/RFC8215/IPv4-compat/6to4 wrappers of an internal IPv4 -\u003e BLOCK; NAT64 of a public IPv4 (`64:ff9b::808:808` = 8.8.8.8) -\u003e ALLOW.\n\n## Impact\n\nA caller who can upload a markdown or textpack document (textpack uploads route to the native engine with zero config; markdown reaches it when the native engine is selected) can make the LightRAG server issue HTTP requests to internal-only addresses that the SSRF guard is meant to forbid, and the fetched body is ingested. On a deployment with NAT64/DNS64 routing this reaches cloud instance metadata (`169.254.169.254`, `100.100.100.200`), loopback services, and RFC1918 internal hosts, exposing IAM credentials and internal service data.\n\n## Suggested fix\n\nDecode the embedded IPv4 of any IPv6 transition wrapper and classify by the embedded IPv4 before the `is_global` check:\n\n```python\ndef _unwrap_embedded_ipv4(ip):\n    \"\"\"Return the IPv4 embedded in an IPv6 transition wrapper (IPv4-mapped,\n    IPv4-compatible, NAT64 64:ff9b::/96 and RFC8215 64:ff9b:1::/48, 6to4\n    2002::/16) so it is classified by IPv4 rules; else return ip unchanged.\"\"\"\n    if ip.version != 6:\n        return ip\n    if ip.ipv4_mapped is not None:\n        return ip.ipv4_mapped\n    if getattr(ip, \"sixtofour\", None) is not None:\n        return ip.sixtofour\n    b = ip.packed\n    if b[:12] == b\"\\x00\" * 12 and b[12:] not in (b\"\\x00\\x00\\x00\\x00\", b\"\\x00\\x00\\x00\\x01\"):\n        return ip_address(b[12:])\n    if b[:12] == b\"\\x00\\x64\\xff\\x9b\" + b\"\\x00\" * 8:\n        return ip_address(b[12:])\n    if b[:6] == b\"\\x00\\x64\\xff\\x9b\\x00\\x01\":\n        return ip_address(b[12:])\n    return ip\n```\n\nThen in `_validated_addresses`:\n\n```python\njudged = _unwrap_embedded_ipv4(ip)\nif not (judged.is_global or any(ip in net for net in allow)):\n    return []\n```\n\nVerified: this blocks the NAT64/IPv4-compat/6to4 internal forms while genuine public IPs and public NAT64-wrapped IPs still pass.\n\n## Resolution\n\nAddressed by [#3426](https://github.com/HKUDS/LightRAG/pull/3426) (pending merge), targeted for release in `lightrag-hku` 1.5.5. `_validated_addresses()` now decides global-routability by **starting from the stdlib\u0027s `is_global` on the literal and only ever tightening it** \u2014 the guard is never more permissive than `ipaddress` itself:\n\n- **Fixed-position transition wrappers are decoded and the embedded IPv4 must also be global** (IPv4-mapped `::ffff:0:0/96`, IPv4-compatible `::/96`, and the NAT64 well-known prefix `64:ff9b::/96` \u2014 RFC 6052 \u00a72.2, contiguous low-32 embedding). So `64:ff9b::7f00:1` (\u2192 loopback) is blocked while `64:ff9b::808:808` (\u2192 public 8.8.8.8) still passes.\n- **Blocks whose embedded-IPv4 position must not be trusted are default-denied as whole blocks**, independent of the interpreter: the RFC 8215 local-use prefix `64:ff9b:1::/48` (technology-agnostic, embedded-IPv4 position not guaranteed per RFC 8215 \u00a75 \u2014 a fixed-offset decode is itself bypassable, e.g. `64:ff9b:1:7f00:0:100:808:808` encodes `127.0.0.1` but its low 32 bits read as public `8.8.8.8`) and 6to4 `2002::/16` (IANA global-reachability N/A; current CPython classifies the whole block non-global \u2014 RFC 7526 deprecated only the 6to4 anycast relay path, not the 2002::/16 prefix itself). Decoding 6to4 by its embedded IPv4 would have re-permitted `2002::/16` addresses the stdlib already blocks, so it is denied outright instead.\n\nGenuine global IPv6 and a NAT64 wrapper of a *public* IPv4 still pass; regression tests cover every wrapped form of loopback / RFC1918 / metadata, the RFC 6052 `/48` suffix bypass, and 6to4 of a public IPv4. Boundary: a NAT64 deployment using a *custom*, globally-routable Network-Specific Prefix (RFC 6052 permits any /32../96), or a genuine local NAT64 / legacy 6to4 in the force-denied blocks, is not auto-permitted \u2014 such operators should pin the download egress or set `NATIVE_MD_IMAGE_ALLOWED_NON_PUBLIC_CIDRS` deliberately.\n\n## Credit\n\ntonghuaroot (tonghuaroot@gmail.com).",
  "id": "GHSA-vv3m-f8x4-7377",
  "modified": "2026-09-22T20:40:25Z",
  "published": "2026-09-22T20:40:25Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/security/advisories/GHSA-vv3m-f8x4-7377"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-85740"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/pull/3426"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/commit/9207e7fd10cf6df3cf26e3dd1921490b34a2f132"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/commit/a25862177a9b00e0edb870778a42b4155924de09"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/commit/c598545aa0084a427bcb19a84ade5e0ec31fa673"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/HKUDS/LightRAG"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/releases/tag/v1.5.5"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "lightrag-hku: SSRF via IPv6-transition address bypass (NAT64, IPv4-compatible, 6to4) of the native-markdown image-download guard"
}

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "lightrag-hku",
        "purl": "pkg:pypi/lightrag-hku"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.5.5"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ],
      "versions": [
        "0.0.2",
        "0.0.3",
        "0.0.4",
        "0.0.5",
        "0.0.6",
        "0.0.7",
        "0.0.8",
        "0.0.9",
        "1.0.0",
        "1.0.1",
        "1.0.3",
        "1.0.5",
        "1.0.6",
        "1.0.8",
        "1.0.9",
        "1.1.0",
        "1.1.1",
        "1.1.2",
        "1.1.3",
        "1.1.4",
        "1.1.5",
        "1.1.6",
        "1.1.7",
        "1.2.1",
        "1.2.2",
        "1.2.3",
        "1.2.5",
        "1.2.6",
        "1.3.0",
        "1.3.1",
        "1.3.2",
        "1.3.3",
        "1.3.4",
        "1.3.5",
        "1.3.6",
        "1.3.7",
        "1.3.8",
        "1.3.9",
        "1.4.0",
        "1.4.1",
        "1.4.10",
        "1.4.11",
        "1.4.11rc2",
        "1.4.12",
        "1.4.12rc1",
        "1.4.13",
        "1.4.13rc1",
        "1.4.14",
        "1.4.15",
        "1.4.16",
        "1.4.2",
        "1.4.3",
        "1.4.4",
        "1.4.5",
        "1.4.6",
        "1.4.7",
        "1.4.8.1",
        "1.4.8.2",
        "1.4.8rc4",
        "1.4.8rc6",
        "1.4.8rc7",
        "1.4.8rc8",
        "1.4.8rc9",
        "1.4.9",
        "1.4.9.1",
        "1.4.9.10",
        "1.4.9.11",
        "1.4.9.2",
        "1.4.9.3",
        "1.4.9.4",
        "1.4.9.4rc1",
        "1.4.9.5",
        "1.4.9.6",
        "1.4.9.7",
        "1.4.9.8",
        "1.4.9.9",
        "1.4.9rc1",
        "1.4.9rc2",
        "1.4.9rc3",
        "1.4.9rc4",
        "1.5.0",
        "1.5.0rc1",
        "1.5.0rc2",
        "1.5.0rc3",
        "1.5.1",
        "1.5.2",
        "1.5.3",
        "1.5.4",
        "1.5.5rc1"
      ]
    }
  ],
  "aliases": [
    "CVE-2026-85740",
    "GHSA-vv3m-f8x4-7377"
  ],
  "details": "## Summary\n\nLightRAG\u0027s native markdown parser downloads external images referenced by an uploaded markdown or textpack document. The only SSRF guard, `_validated_addresses()` in `lightrag/parser/markdown/parser.py`, resolves the image host and rejects it when the resolved IP is not `is_global`. That check is evaluated on the raw resolved address and never decodes IPv6 transition wrappers that embed an internal IPv4. Python\u0027s `ipaddress` classifies a NAT64 (`64:ff9b::/96` and the RFC 8215 `64:ff9b:1::/48` prefix), IPv4-compatible (`::a.b.c.d`), or 6to4 (`2002::/16`) literal that wraps an internal IPv4 as globally routable, so the guard passes it. On a host with NAT64/DNS64 routing the request is then delivered to the embedded internal target (loopback, RFC1918, or a cloud metadata endpoint), and the fetched body is ingested. The plain and IPv4-mapped (`::ffff:`) forms of the same internal address are correctly blocked, so this is an encoding that defeats the existing filter.\n\n## Affected component and versions\n\n- Package: `lightrag-hku` (LightRAG), the native markdown image-download path.\n- Component: `lightrag/parser/markdown/parser.py`, guard `_validated_addresses()` (the `if not (ip.is_global or ...)` check), reached from `_download()` -\u003e `_build_guarded_opener().open(req)`.\n- Enabled by default: `download_enabled = _env_bool(\"NATIVE_MD_IMAGE_DOWNLOAD_ENABLED\", True)`.\n- Affected: LightRAG `\u003c= 1.5.4` (latest release at time of report, commit `9a45b64`).\n- Precondition: the caller can upload a document (API key via `Depends(combined_auth)`); the host has NAT64/DNS64 routing for the encoded address to reach the internal endpoint.\n\n## Vulnerable code vs the guarded sibling\n\nEvaluated on the exact resolved addresses (CPython `ipaddress`):\n\n| Form | address | is_global | guard verdict |\n|---|---|---|---|\n| plain internal 127.0.0.1 | `127.0.0.1` | False | BLOCK (correct) |\n| IPv4-mapped | `::ffff:7f00:1` | False | BLOCK (correct) |\n| NAT64 64:ff9b::/96 | `64:ff9b::7f00:1` | True | PASS (bypass) |\n| NAT64 RFC8215 64:ff9b:1::/48 | `64:ff9b:1::7f00:1` | True | PASS (bypass) |\n| IPv4-compatible ::a.b.c.d | `::7f00:1` | True | PASS (bypass) |\n| 6to4 2002::/16 | `2002:7f00:1::` | True | PASS (bypass) |\n\nThe guard already blocks the plain and IPv4-mapped internal forms (the forms the author tested), proving the intent is to reject internal destinations; it never canonicalizes the other transition wrappers, so they pass.\n\n\u003e **Note \u2014 CPython version dependency of the table.** The `is_global` verdicts above depend on the interpreter. CPython gh-113171 (backported to 3.10.14 / 3.11.9 / 3.12.3+, ~2024-04) added `64:ff9b:1::/48` and `2002::/16` to `ipaddress._private_networks`, so on any current patch release those two rows return `is_global = False` and are already blocked by the stdlib before the guard is reached. The **`64:ff9b::/96` well-known NAT64 prefix and the IPv4-compatible `::/96` form bypass on every current CPython** \u2014 and `64:ff9b::/96` is exactly the prefix real DNS64/NAT64 deployments use, so the vulnerability stands. The four-row table reflects a pre-gh-113171 interpreter. The fix classifies by the embedded IPv4 and therefore closes all forms regardless of interpreter version.\n\n## Severity\n\nHigh. CWE-918 (Server-Side Request Forgery). CVSS 3.1 vector `AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:L/A:N`, base score 7.1. AC:H reflects the NAT64/DNS64 network precondition for end-to-end reach; PR:L because document upload requires the API key; S:C because the request pivots into an internal network segment; C:H for internal/metadata/secret disclosure.\n\n## Proof of concept (deployed, both directions, verbatim)\n\nEnvironment: three Docker containers on an IPv6-enabled network - an internal victim (`10.66.0.2:80`, RFC1918) serving a unique secret, a DNS64/NAT64 gateway owning `64:ff9b::/96` that forwards `[64:ff9b::0a42:0002]:80` to `10.66.0.2:80`, and an attacker host with a client-side route `64:ff9b::/96 via \u003cgateway\u003e`. This models a client on an IPv6-only / NAT64+DNS64 network (AWS/GCP IPv6-only subnets, many mobile and corporate networks). The attacker runs the EXACT released guard code (`_validated_addresses`, `_host_is_public`, `_build_guarded_opener`, the `_Guarded*Connection/Handler` classes, sliced verbatim from v1.5.4 `parser.py`, only the `logger` import shimmed) plus a verbatim reproduction of the released `_download()` scheme and host precheck. Trigger in production: an uploaded document containing `![x](http://[64:ff9b::\u003cinternal-hex\u003e]/a.png)`.\n\nRUN 1, released v1.5.4 guard (VULNERABLE):\n\n```\nattack URL = http://[64:ff9b::0a42:0002]:80/logo.png\n===== DIRECTION A: ATTACK (NAT64-wrapped internal address) =====\nguard verdict : PASSED (fetched)\nSECRET LEAKED : True\nbody          : SSRF-LIGHTRAG-INTERNAL-METADATA-9f3a2b7c\n===== control: plain internal address (must be BLOCKED) =====\nverdict: BLOCKED at guard: non-public host blocked\n===== DIRECTION B: PUBLIC address (must still be FETCHED) =====\nguard verdict : PASSED (fetched)\npublic body   : PUBLIC-OK-BODY\n===== SUMMARY =====\nattack_leaked_internal_secret = True\nplain_internal_blocked        = True\npublic_still_fetched          = True\n```\n\nRUN 2, one-line-class decode fix applied (same inputs, same network):\n\n```\n===== DIRECTION A: ATTACK (NAT64-wrapped internal address) =====\nguard verdict : BLOCKED at guard: non-public host blocked\nSECRET LEAKED : False\n===== control: plain internal address (must be BLOCKED) =====\nverdict: BLOCKED at guard: non-public host blocked\n===== DIRECTION B: PUBLIC address (must still be FETCHED) =====\nguard verdict : PASSED (fetched)\npublic body   : PUBLIC-OK-BODY\n===== SUMMARY =====\nattack_leaked_internal_secret = False\nplain_internal_blocked        = True\npublic_still_fetched          = True\n```\n\nFix-correctness matrix (isolated): genuine global IPv6 `2606:4700:4700::1111` and `2001:4860:4860::8888` -\u003e ALLOW (no false positive); NAT64/RFC8215/IPv4-compat/6to4 wrappers of an internal IPv4 -\u003e BLOCK; NAT64 of a public IPv4 (`64:ff9b::808:808` = 8.8.8.8) -\u003e ALLOW.\n\n## Impact\n\nA caller who can upload a markdown or textpack document (textpack uploads route to the native engine with zero config; markdown reaches it when the native engine is selected) can make the LightRAG server issue HTTP requests to internal-only addresses that the SSRF guard is meant to forbid, and the fetched body is ingested. On a deployment with NAT64/DNS64 routing this reaches cloud instance metadata (`169.254.169.254`, `100.100.100.200`), loopback services, and RFC1918 internal hosts, exposing IAM credentials and internal service data.\n\n## Suggested fix\n\nDecode the embedded IPv4 of any IPv6 transition wrapper and classify by the embedded IPv4 before the `is_global` check:\n\n```python\ndef _unwrap_embedded_ipv4(ip):\n    \"\"\"Return the IPv4 embedded in an IPv6 transition wrapper (IPv4-mapped,\n    IPv4-compatible, NAT64 64:ff9b::/96 and RFC8215 64:ff9b:1::/48, 6to4\n    2002::/16) so it is classified by IPv4 rules; else return ip unchanged.\"\"\"\n    if ip.version != 6:\n        return ip\n    if ip.ipv4_mapped is not None:\n        return ip.ipv4_mapped\n    if getattr(ip, \"sixtofour\", None) is not None:\n        return ip.sixtofour\n    b = ip.packed\n    if b[:12] == b\"\\x00\" * 12 and b[12:] not in (b\"\\x00\\x00\\x00\\x00\", b\"\\x00\\x00\\x00\\x01\"):\n        return ip_address(b[12:])\n    if b[:12] == b\"\\x00\\x64\\xff\\x9b\" + b\"\\x00\" * 8:\n        return ip_address(b[12:])\n    if b[:6] == b\"\\x00\\x64\\xff\\x9b\\x00\\x01\":\n        return ip_address(b[12:])\n    return ip\n```\n\nThen in `_validated_addresses`:\n\n```python\njudged = _unwrap_embedded_ipv4(ip)\nif not (judged.is_global or any(ip in net for net in allow)):\n    return []\n```\n\nVerified: this blocks the NAT64/IPv4-compat/6to4 internal forms while genuine public IPs and public NAT64-wrapped IPs still pass.\n\n## Resolution\n\nAddressed by [#3426](https://github.com/HKUDS/LightRAG/pull/3426) (pending merge), targeted for release in `lightrag-hku` 1.5.5. `_validated_addresses()` now decides global-routability by **starting from the stdlib\u0027s `is_global` on the literal and only ever tightening it** \u2014 the guard is never more permissive than `ipaddress` itself:\n\n- **Fixed-position transition wrappers are decoded and the embedded IPv4 must also be global** (IPv4-mapped `::ffff:0:0/96`, IPv4-compatible `::/96`, and the NAT64 well-known prefix `64:ff9b::/96` \u2014 RFC 6052 \u00a72.2, contiguous low-32 embedding). So `64:ff9b::7f00:1` (\u2192 loopback) is blocked while `64:ff9b::808:808` (\u2192 public 8.8.8.8) still passes.\n- **Blocks whose embedded-IPv4 position must not be trusted are default-denied as whole blocks**, independent of the interpreter: the RFC 8215 local-use prefix `64:ff9b:1::/48` (technology-agnostic, embedded-IPv4 position not guaranteed per RFC 8215 \u00a75 \u2014 a fixed-offset decode is itself bypassable, e.g. `64:ff9b:1:7f00:0:100:808:808` encodes `127.0.0.1` but its low 32 bits read as public `8.8.8.8`) and 6to4 `2002::/16` (IANA global-reachability N/A; current CPython classifies the whole block non-global \u2014 RFC 7526 deprecated only the 6to4 anycast relay path, not the 2002::/16 prefix itself). Decoding 6to4 by its embedded IPv4 would have re-permitted `2002::/16` addresses the stdlib already blocks, so it is denied outright instead.\n\nGenuine global IPv6 and a NAT64 wrapper of a *public* IPv4 still pass; regression tests cover every wrapped form of loopback / RFC1918 / metadata, the RFC 6052 `/48` suffix bypass, and 6to4 of a public IPv4. Boundary: a NAT64 deployment using a *custom*, globally-routable Network-Specific Prefix (RFC 6052 permits any /32../96), or a genuine local NAT64 / legacy 6to4 in the force-denied blocks, is not auto-permitted \u2014 such operators should pin the download egress or set `NATIVE_MD_IMAGE_ALLOWED_NON_PUBLIC_CIDRS` deliberately.\n\n## Credit\n\ntonghuaroot (tonghuaroot@gmail.com).",
  "id": "PYSEC-2026-4064",
  "modified": "2026-10-01T17:10:27.401407Z",
  "published": "2026-10-01T16:38:37.571208Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/security/advisories/GHSA-vv3m-f8x4-7377"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-85740"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/pull/3426"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/commit/9207e7fd10cf6df3cf26e3dd1921490b34a2f132"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/commit/a25862177a9b00e0edb870778a42b4155924de09"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/commit/c598545aa0084a427bcb19a84ade5e0ec31fa673"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/HKUDS/LightRAG"
    },
    {
      "type": "WEB",
      "url": "https://github.com/HKUDS/LightRAG/releases/tag/v1.5.5"
    },
    {
      "type": "PACKAGE",
      "url": "https://pypi.org/project/lightrag-hku"
    },
    {
      "type": "ADVISORY",
      "url": "https://github.com/advisories/GHSA-vv3m-f8x4-7377"
    }
  ],
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "lightrag-hku: SSRF via IPv6-transition address bypass (NAT64, IPv4-compatible, 6to4) of the native-markdown image-download guard"
}